We aim to develop, test, and apply a drastically new computational methodology for the analysis of more than one complex phenotype at a time, with the goal of generating novel biological results. Specifically, we propose to design and validate a battery of novel analytical tools for the inference of causal relationships among human genomic variations, environmental factors, and more than one mental health phenotype, explicitly exploiting the genetic and environmental non-independence of complex (multigenic) disorders. PUBLIC HEALTH RELEVANCE: We attempt to consolidate in a single modeling framework a number of disparate approaches for analysis of complex neuropsychiatric disorders. The comprehensive modeling approach will produce experimentally testable predictions, a considerable number of which we will be able to validate within the proposed research. We will focus on several phenotypes with major impacts on the health of US populations, such as anxiety, schizophrenia and depression.